silencethendang

It seems like Einstein was saying "there's no such thing as a theory-free fact," specifically in his case that observation itself is influenced by some pre-existing theory. Some deduction is involved. But this begs the question... where did *that* theory originate?

I said something similar in my response. I think your first point depends on how much reliance on LLM's determines things like quality of writing and general creativity. For now, comedy (which is a big source of reposted content) probably will always come from humans. To the extent that LLM's have invaded comedy, it's from humans using them in funny ways and reposting that.

LLM's need to be trained on something, and the more the 'market' for text is flooded with ChatGPT output, the more relatively scarce human-written text will be. However, if in the long-term, LLM's hinder peoples' ability to string together sentences or otherwise affect the quality of their writing, then (modern) human-written text might not be so valuable. In this case, the value of past human-written text will skyrocket (Twitter, Reddit, etc. will make a killing).

The complexity exists because it creates demand for products that eventually become entrenched over time (Terraform, Kubernetes, DevOps products, SCRUM/Agile etc.). I don't think this is necessarily bad, but I don't think the complexity will ever stop existing until the incentive structure to keep it in place goes away. It's hard to reverse complexity, so I can't really see this happening.

I do think that the definition of 'growth' will change from spending money (i.e. 'growing a team' or 'growing headcount') and sheer output (we hit X goals this year) to actual long-term profitability (not just EPS like it's been for the last year or so). I don't personally benefit from this (since I am a tech worker) but from a social standpoint it's probably a good thing.

These articles make me question what "success" means in the AI space. Is "success" just making a lot of money or raising a giant funding round or achieving a valuation? If so, then yeah, there's probably some room for at least a few startups to get there by perfectly the surfing the hype wave. I mean, so many are trying, at least one is bound to succeed by pure chance, even if nothing they're doing is particularly groundbreaking... i.e. the "wrapper around ChatGPT" business model that happens to get sold to just the right executives or PE guys or whatever.

However, if by "success" we mean truly revolutionary technology, I find it difficult to think that this will actually come out of any company trying to be profitable in the current environment. There's just too much urgency around things that are concrete, have low uncertainty, and immediate payoff, either in terms of checked boxes engineering managers can report upwards or 'initiatives' that executives can show their boards. I can't imagine doing anything resembling 'research' in that kind of environment. We'll probably have to wait until after the next recession for companies to get comfortable with throwing money and headcount at these kinds of things.

Sorry for the late reply. I've never heard of the Al Roth book, but seems interesting. I'll check it out. I mostly became familiar with Mechanism Design when I took the first year microeconomics sequence in graduate school (basically one of the few departments this stuff is covered academically is in Economics). So my best reference that I can vouch for is, unfortunately, probably this: https://global.oup.com/ushe/product/microeconomic-theory-978...

There's also Matthew Jackson's book which might be better. I've read his class notes for other topics and I think he does a good job explaining: https://web.stanford.edu/~jacksonm/mechtheo.pdf

For #2 - I'm not a PE insider, but I'm pretty aware of a lot of corporate dynamics in big tech, which increasingly seems like it's indistinguishable from any other corporate sector. I think we have a habit of thinking of companies/firms as individual actors deciding what's in their best interest, when in reality it's basically an oligarchy of either C-suite execs or boards of directors determining what's going to happen to the rest of the company. If these people want to get rich quick, and I'm guessing they do, since this is probably the end game for 99% of them, then private equity is their answer. They likely don't really care about the employees or the future of the company as an end in itself, just a means to enrich themselves, and so if PE presents this opportunity without those strings attached (i.e. by splitting the dividend recapitalization or the real estate leasebacks with the execs in the form of 'bonuses') the execs/boards will likely take it.

TL;DR - PE can't exist in its current form without extreme power/ownership imbalances within the firms they take over and run into the ground. This has probably been increasing over time, hence why PE has been getting worse over time.

I know aligning incentives is usually something that falls under the broad category of "management" or people skills in a company, but there's a subfield of economics that studies exactly how to do this in a structured way (https://en.wikipedia.org/wiki/Mechanism_design). It's usually applied to things like auctions, but I've always thought applying incentive alignment in a systematic way (in which, for instance, middle management submits "bids" for metrics that they want to track and for headcount) could be a viable way to run things. Any thoughts on something like this?

I've been reading this book. I’m not actually done with it, but I’ve gotten through a pretty substantial portion of it. The author clearly has a bias, having worked as the DOJ as a prosecutor, but also has the facts to back up what he’s saying. It’s not like this bias makes him not credible. However, there were times when reading I would say to myself, “yeah but a company would probably justify things this way…” I’ve seen firsthand a lot of rationalization of things that seem bad, so maybe this is just how I am now.

My takeaways thus far:

1. The author seems to put the blame squarely on the shoulders of private equity firms. While they’re obviously not blameless, a lot of the blame should also go toward the C-suite executives and the boards that run the companies that are eventually bought by these firms. I’m guessing that for the most part, these schemes, like taking out massive debt to pay the PE firm, also hugely benefits the executives that are probably agreeing to sign over the firm for a payday. If I’m understanding ‘dividend recapitalization’ correctly, it’s unsurprising that an executive at a target company would turn down a “loan” that they themselves literally never have to pay back. I mean, corporate executives are probably the kind of people that have been waiting their whole lives for someone to make them this exact offer.

2. I’m getting to the part on how some of these cases have been prosecuted, but it doesn’t seem like the fines for doing anything illicit ever outweigh the benefit of doing so in purely dollar terms… as the author says, the lawsuits and fines faced by these firms are “just a cost of doing business.” If the government is getting paid the fine out of the profits made by illicit activity, then this pretty much fits the definition of extortion by the government as David Graeber has pointed out (somewhere, I don’t remember where). PE is definitely not the only industry that this occurs in but I guess I wasn’t aware of how entrenched they are, especially with the GOP.

3. The easiest way of thinking about these PE firms is almost as if they weren’t run by humans but as if someone wrote a program to extract profits over a 5-year horizon by any means possible. You almost have to admire some of the legal loopholes that they’re able to exploit (and question why they exist). Usually there is at least some potentially fake virtual signaling that comes with “public” capitalism, but in the PE world, it seems like none of that exists (with the pretty hilarious exception of Tom Gores, who released this statement while benefitting from charges from prison phone calls). Most “steady revenue streams” and things that make sense on paper AND abide by the social contract are probably already owned, and anything left screws someone over, somewhere in a pretty unethical way.

4. Following from the last two points, my guess is that the thing that PE firms fear most is press and journalism, not because it affects them directly (they probably don’t care about that) but because it puts pressure on companies and funds and whoever else is involved with them to divest if they don’t want to be considered evil. Like I said before, profits that look good on paper and in Excel are almost undoubtedly screwing someone over somewhere, and people (read: executives) want these profits, but they don’t want to know how they’re obtained. PE firms do this dirty work. A big, underrated selling point of PE seems to be blame avoidance – if you’re the CEO at a major retailer, you hate your job, and you want to retire in Beverly Hills now with absolutely no mortgage payments, then PE is your answer. If someone starts attacking you on Twitter when the benefits get cut, you’ve got a great alibi – “I had no idea Blackstone was gonna do this!”